In an image-capturing device in which image-capturing elements are arranged in a two-dimensional matrix form, a technique for obtaining multiple pieces of image data by performing image-capturing process multiple times and combining the multiple pieces of image data is well known as indicated in, for example, JP 2000-244797 A. Alternatively, in order to cope with an image-capturing scene having great difference in the brightness, N pieces of image data of which sensitivities are different are obtained by performing image-capturing process of a still picture N times (N is equal to or larger than 2) with different exposure times, and an HDR-combined image is obtained by combining these N pieces of image data. However, in the method for obtaining these N pieces of image data, it takes a certain period of time to obtain the N pieces of image data, and the simultaneousness cannot be ensured, and in addition, artifact occurs due to the combining of the N pieces of image data, and it is difficult to cope with motion pictures.
A technique for arranging image-capturing elements of different sensitivities adjacent to each other, and expanding the dynamic range by combining a signal from an image-capturing element of a higher sensitivity and a signal from an image-capturing element of a lower sensitivity is well known as indicated in JP 2006-270364 A. In a technique disclosed in JP 2006-253876 A, the difference in the sensitivity is given within one piece of image data by performing control so as to change the electrical-charge accumulation time of each image-capturing element for each area of the image-capturing unit by controlling the shutter speed.